Clinical investigations
Radioimmunoguided imaging of prostate cancer foci with histopathological correlation

https://doi.org/10.1016/S0360-3016(00)01582-0Get rights and content

Abstract

Purpose: We have previously presented a technique that fuses ProstaScint and pelvic CT images for the purpose of designing brachytherapy that targets areas at high risk for treatment failure. We now correlate areas of increased intensity seen on ProstaScint-CT fusion images to biopsy results in a series of 7 patients to evaluate the accuracy of this technique in localizing intraprostatic disease.

Methods and Materials: The 7 patients included in this study were evaluated between June 1998 and March 29, 1999 at Metrohealth Medical Center and University Hospitals of Cleveland in Cleveland, Ohio. ProstaScint and CT scans of each patient were obtained before transperineal biopsy and seed implantation. Each patient’s prostate gland was biopsied at 12 separate sites determined independently of Prostascint-CT scan results.

Results: When correlated with biopsy results, our method yielded an overall accuracy of 80%: with a sensitivity of 79%, a specificity of 80%, a positive predictive value of 68%, and a negative predictive value of 88%.

Conclusion: The image fusion of the pelvic CT scan and ProstaScint scan helped identify foci of adenocarcinoma within the prostate that correlated well with biopsy results. These data may be useful to escalate doses in regions containing tumor by either high-dose rate or low-dose rate brachytherapy, as well as by external beam techniques such as intensity modulated radiotherapy (IMRT).

Introduction

Prostate brachytherapy is a continuously evolving technique for the treatment of localized prostate adenocarcinoma. Early attempts at brachytherapy with permanent radioisotopes in the 1970s were marked by unsatisfactory results due to technology inadequate for proper treatment planning. The widespread availability of improved imaging modalities and planning software in the 1980s led to the development of prostate brachytherapy as it is widely practiced today, with a low-morbidity transperineal approach that can be performed on an outpatient basis (1). Modern prostate brachytherapy relies on computed tomography (CT) or transrectal ultrasound (TRUS)-based calculations of prostate size and shape to design a seed distribution that provides the prescribed dose of radiation to the volume of the gland while minimizing exposure to the rectum and urethra. This approach is based primarily on the geometry of the prostate and does not take into account the distribution of tumor within the gland. Theoretically, therapy designed to target specific foci of disease within the prostate would yield better local disease control with decreased morbidity.

One of the major impediments to this approach has been the lack of modalities to accurately detect and localize intraprostatic disease. Techniques such as ultrasound, CT, or magnetic resonance imaging (MRI) that are traditionally used to image prostate cancer lack the sensitivity and/or specificity needed to assess intraprostatic disease effectively. TRUS is useful for assessing organ volume but fails to detect up to 40% of cancers that appear isoechoic (2). Another ultrasound based study, color doppler sonography (CDS), has a low sensitivity (49%) for detecting prostate cancer (3). When used together, TRUS and CDS appear to demonstrate a positive predictive value as high as 77%, but at the cost of a diminished sensitivity (2). CT scans can evaluate prostate volume but offer staging ability comparable only to digital rectal examination (DRE) 2, 4. Of the imaging modalities widely available, MRI appears to be the most useful. However, its use is typically limited to the evaluation of extracapsular and seminal vesicle invasion in the context of selecting patients for prostatectomy (5). It is effective in detecting peripheral gland tissue variations, but has a poor positive predictive value due to its inability to distinguish the reduced signal of adenocarcinoma from that of hormone therapy, scarring, or prostatitis. It also fails to detect the 30% of tumors that occur within the central gland and demonstrates a decreased sensitivity for lesions less than 5 mm in size 6, 7. Magnetic resonance spectroscopy (MRS) relies on the metabolic activity of cancers and appears to boost the specificity of endorectal coil MRI (8). Its use in localizing intraprostatic disease for therapeutic strategy is currently being explored at other institutions 9, 10, 11, 12. We present a technique for assessing intraprostatic disease that utilizes ProstaScint nuclear medicine scans and CT images.

ProstaScint is an Indium-111 labeled monoclonal mouse antibody (CYT-356 or capromab pendetide) specific for prostate-specific membrane antigen (PSMA). It is FDA-approved for radioimmunoscintographic detection of disease recurrence in post-treatment patients with rising PSAs and for detecting metastatic disease in pretreatment high-risk patients (13). One feature that potentially limits the specificity of ProstaScint scans is the uptake of antibody in the vasculature, leading to increased signal in muscle, blood vessels, and bones. Our method utilizes these background signals as anatomic markers to correlate ProstaScint scans with CT scans of the pelvis (14). This allows coregistration of signals seen on the nuclear scan with anatomic structures seen on CT scan.

Patients in our study initially received a ProstaScint scan as part of their routine work-up to assess candidacy for brachytherapy. Additionally, all patients had a pelvic CT scan for image fusion and to determine prostate volume or rule out pubic arch obstruction. Once these graphic data were collected, they were coregistered to generate fusion images that superimposed the radioimmunoscintographic signals of the ProstaScint scan onto the anatomy of the CT scans. We have previously discussed the feasibility of this technique and its use in directing brachytherapy toward cancer foci within the prostate (14).

Although ProstaScint imaging has been criticized for having false positives beyond the prostate gland when used for evaluation of metastatic disease, we believe that we have overcome some of the initial difficulties in interpreting the scans through the use of improved imaging techniques, such as an increased matrix of 128 × 128 during imaging, and iterative reconstruction to improve resolution of the study. Additionally, coregistration to CT or MRI has greatly improved the ability to interpret regions of increased antibody uptake in the pelvis, abdomen, or chest by determining the underlying anatomic structure such as small bowel or vascular regions which are prone to producing false positives. We believe that the ProstaScint scan may be useful for all patients undergoing prostate brachytherapy, as well as possibly for patients treated with external beam radiation using intensity modulated radiation therapy (IMRT). Not only does the ProstaScint scan allow assessment of possible metastatic disease, as it is routinely used, but also allows for assessment of intraprostatic tumor foci through image fusion with either CT scan or MRI to provide anatomic visualization of the prostate gland.

The SPECT images can be readily coregistered to the CT scan or MRI through matching of internal vascular structures and regions containing bone marrow, which have high concentrations of the monclonal antibody. By aligning the two image studies in multiple planes throughout the pelvis, one can be reasonably confident that the coregistration within the prostate gland is accurate. The CT scan is obtained immediately following completion of the SPECT study, without allowing the patient to use the restroom between scans, thus minimizing prostate movement between the studies. Once the coregistration is completed, the contrast is increased on the SPECT images to remove any residual uptake seen outside of the prostate within the surrounding musculature. Residual activity may be seen within the vascular and marrow-containing structures. Regions within the prostate that persist to have visualized antibody concentration are selected as target lesions for radiotherapy either by brachytherapy seed placement within the region, or possibly by IMRT targeted dose escalation.

While MRS may also be able to help identify regions within the prostate for targeting with either brachytherapy or IMRT, it would be difficult to access the pelvis or abdomen due to the amount of time required to scan such a large area by spectroscopy. We feel that the use of radioimmunoguided prostate brachytherapy has a possible advantage over MRS-guided prostate brachytherapy, due to its ability to routinely evaluate for the risk of extraprostatic disease. When confirmed by biopsy, this would help select out patients for whom brachytherapy would not be indicated. Currently, patients who have extraprostatic antibody concentration detected in regions not easily confirmed by biopsy, such as the para-aortic region, are assumed to have a false positive and are treated with local radiotherapy and followed for disease progression. While false positives have been recorded in approximately 10% of all of our brachytherapy patients to date at various regions in either the chest or abdomen that could not be confirmed by biopsy, none of these patients has yet demonstrated a clinical or biochemical failure with limited follow-up. We realize the difficulty in interpreting these extraprostatic false positives, but believe that the more useful aspect of the antibody image may be within the gland itself by allowing the use of regional dose escalation within the high-risk regions of the prostate gland. Prostate-specific membrane antigen (PSMA) is upregulated and overexpressed by both cancerous tissue and prostatic intraepithelial neoplasia (PIN), but not atypical adenomatous hyperplastic lesions (AAH.) or benign prostatic hypertrophy (BPH) (15). While the expression of PSMA in PIN may result in a false positive reading for carcinoma within the prostate gland with the image fusion, we believe that PIN should be targeted and treated as cancer during brachytherapy or IMRT, as it has been shown to be a likely precursor of invasive carcinoma (16). We now present biopsy data from 7 patients to assess the accuracy of this imaging technique.

Section snippets

Methods and materials

Between June 1998 and March 1999, 7 patients with clinical Stage II (5 T1c, 2 T2a) prostate adenocarcinoma were treated with ultrasound-guided transperineal implantation of I-125 or Pd-103 at MetroHealth Medical Center, Cleveland, Ohio, and University Hospitals of Cleveland. Pretreatment patient data are listed in Table 1. Preoperative evaluation included a complete history and physical examination, a ProstaScint scan, a thin-slice 5-mm pelvic CT scan with oral and IV contrast, prostate volume

Results

Overall, for the 84 biopsies obtained from these 7 patients, we had 23 true positive (TP), 44 true negative (TN), 11 false positive (FP), and 6 false negative (FN) readings. There did not appear to be a strong correlation between anatomic site and TP/TN/FP/FN rate. Of note, the LAB, LPM, and LAA had a high TN rate and a low TP rate. When the anterior portion of the gland was compared with the posterior, the TP/TN/FP/FN rates were very similar. The anterior gland had 10 TP, 22 TN, 6 FP, and 4 FN

Discussion

The goal of our current study is to provide a prospective correlation between the fusion study images and histopathology specimens. Our results are encouraging. Our sensitivity of 79% and specificity of 80% surpass values reported for current FDA-approved ProstaScint uses (20). Our high NPV of 88% also suggest that it may be safe to decrease radiation exposure to sensitive tissues, such as the rectum or urethra, without undue fear of failure to treat undetected disease. The anterior base of the

Cited by (50)

  • A Prospective Trial of Intensity Modulated Radiation Therapy (IMRT) Incorporating a Simultaneous Integrated Boost for Prostate Cancer: Long-term Outcomes Compared With Standard Image Guided IMRT

    2017, International Journal of Radiation Oncology Biology Physics
    Citation Excerpt :

    ProstaScint (Cytogen Corp, the manufacturer of indium-111-capromab pendetide. ProstaScintTM is currently owned by Rosewind Corporation) can be used to visualize an intraprostatic lesion (IPL) that can be boosted during intensity modulated radiation therapy (IMRT) (4). This prospective study assessed the feasibility and outcome of prostate IMRT delivering a ProstaScint guided simultaneous integrated boost (SIB) to an IPL.

  • Detection and localization of carcinoma within the prostate using high resolution transrectal gamma imaging (TRGI) of monoclonal antibody directed at prostate specific membrane antigen (PSMA) - Proof of concept and initial imaging results

    2013, European Journal of Radiology
    Citation Excerpt :

    Focal areas of uptake were able to be demonstrated in a prostatic quadrant with biopsy-proven malignancy in a majority of patients. When the threshold for a positive finding on TRGI was set at a target-to-background ratio of 2.0 or 2.5, the sensitivity, specificity, PPV, NPV, and accuracy in detecting malignancy were relatively modest and similar to those reported in the early literature for In-111 capromab pendetide with planar scintigraphy in the detection of nodal metastases from prostate cancer or recurrence in the prostatic fossa [5,10]. Given the crude localization of uptake on TRGI without a corresponding anatomic imaging correlate, the known limitations of In-111 capromab pendetide as an imaging agent, and the other limitations involving study design described below, the results of the ability of TRGI to specifically detect foci of prostate cancer in vivo are only very preliminary.

  • Prostate-specific membrane antigen-based imaging

    2013, Urologic Oncology: Seminars and Original Investigations
    Citation Excerpt :

    Sodee et al. suggested that with experience in over 600 cases and a detailed case report of 5 patients, this technique is likely to improve the high false negative rate, but there is scant pathology proven evidence to the contrary [47]. Using the fusion techniques, Ellis et al. have reported a sensitivity of 79% and specificity of 80% with the capromab-CT [48]. While no formal prostate imaging studies of humanized J591 have been conducted, 2 independent phase I therapeutic trials have been completed where imaging was performed.

  • Prostate Cancer. Anatomy and Histology.

    2012, Oncologic Imaging: A Multidisciplinary Approach Expert Consult
View all citing articles on Scopus
View full text